Effects of Briquette Blend on Packing Structure of Fine Coal Portion

Abstract

Briquette blending aims to increase coke strength by increasing the bulk density of the coal charge by blending in high-density briquettes. This technique tends to decrease the bulk density of the powder coal portion of the coal blend. In this study, we attempted to elucidate the mechanism behind this decrease in the bulk density of the powder coal portion owing to briquette blending. We conducted drop tests of blended coal that included briquettes, observed the behaviour of the dropped coal using a high-speed camera, and then performed quantitative estimation of the change in bulk density by observing the resulting coal packing structure using an X-ray CT system and image analysis. The bulk density of the powder coal portion decreased owing to the formation of localised low-bulk-density regions around the briquettes. Two types of low-bulk-density regions exist. In the first case, the scattering of the powder coal by the impact of the falling briquette forms gaps, which remain in the form of large voids under the briquettes after charging. The second type is presumably due to the large difference in size between the briquette and powder coal, which causes a wall effect between them. We then used a newly developed image analysis process to estimate the widths of the two types of low-bulk-density regions quantitatively as 8–10 mm and 5–6 mm. This study demonstrated how briquette blending creates anisotropic low-bulk-density regions around the briquettes, which leads to a decrease in the bulk density of the powder coal portion.

1. Introduction

In the ironmaking process using a blast furnace, coke quality has a significant impact on the operational stability and reducing agent ratio. Coke acts as a reducing agent, heat source, and spacer to secure the flow path of gas and liquid in the blast furnace. Its role as a spacer is particularly important, and the coke breakage behaviour in the blast furnace has a significant impact on the operation of the blast furnace. Therefore, various pre-treatment techniques have been developed to produce coke with high strength.

One of the operating conditions that affects coke strength is the bulk density of the coal charge. Nomura et al.¹⁾ observed that coal particles should be sufficiently adhered and bonded to each other to obtain high-strength coke, and the factors affecting coal bonding are the bulk density of the initial coal charge and the dilatability of the coal. With regard to coal pre-treatment facilities, to increase the bulk density of the coal charge, various technologies have been developed, including the coal moisture control process²⁾ in which coal is dried to adjust coal moisture before being charged into the coke oven, and a stamping process³⁾ in which coal is preformed into a dense cake before being charged into the coke oven.

One such coal pre-treatment method aimed at improving the bulk density of the coal charge is the briquette blending method,⁴⁾ in which a portion of the powder coal is briquetted and then the briquettes are charged into the coke oven together with the rest of the powder coal. As a briquette has higher apparent density than the powder coal, briquette blending increases the overall bulk density of the coal charge, thereby improving the coke strength.

On the other hand, Okuhara et al.⁵⁾ note that the bulk density of the powder coal portion decreases with briquette blending. Figure 1 shows the relationship between the briquette blend ratio and the bulk density of the coal charge. The figure shows that the bulk density of the coal charge increases as the briquette blend ratio increases up to a briquette ratio of 70%. On the other hand, these actual measurements are lower than the values indicated by alternate long and short dashed lines, which represent the estimated values of the coal bulk density calculated as the weighted average of the apparent density of the briquettes and the bulk density of the powder coal without briquettes. As the apparent density of the briquette does not change, the difference between the lines must be due to the decrease in the density of the powder coal portion. Based on the assumption, the estimated bulk density of the powder coal portion should change as the dashed line in the figure. As described above, bulk density is a major factor that influences coke strength, and the decrease in bulk density of the powder coal portion could lower the coke strength. Okuhara et al.⁵⁾ indicated that blending briquettes with lower dilatation ability resulted in lower coke strength than blending no briquettes.

Fig. 1.

Change in the bulk density of charged coal in a coke chamber by briquette blend.⁵⁾ (Briquette and powder coal (measured): measured coal bulk density in a coke oven chamber, Briquette and powder coal (estimated): estimated value of coal bulk density in a coke oven chamber calculated using the apparent density of briquettes and the bulk density of powder coal without briquettes, powder coal portion (calculated): bulk density of powder coal portion calculated based on the change in the coal bulk density in the chamber and the apparent density of briquettes).

There have been studies of packing density and other properties in particulate binary mixtures. Ricardo et al.⁶⁾ mentioned that wedging and wall effect caused deviation of porosity from a theoretical model in binary mixtures with significant size ratios. However, the phenomenon that causes the bulk density of the powder coal portion to decrease owing to briquette blending did not directly confirmed. Therefore, in this study, we report the results of using an X-ray CT system to estimate the changes in the packing structure of the powder coal portion owing to briquette blending.

2. Test Method

We used drop testing to simulate the packing structure of the coal, and then used an X-ray CT system to observe the packing structure of the coal. We then applied image analysis to the obtained X-ray CT images to visualise the 3D structure of the coal and estimate the changes in the packing density of the powder coal portion quantitatively.

2.1. Coal Drop Test

To prepare the powder coal, we pulverised coal blend until the ratio under 3 mm was 82.5% to 85% (mean particle size: 1.6 mm to 1.8 mm) and then adjusted the moisture content to 10%. For the briquettes, we used pillow-shaped (cup size: 55 W×40 L×23 D mm), 30 cc, +30 mm briquettes, which is more than 15 times larger than the particles composing the powder coal portion. For the drop test, we used a drop device approximately one-half the size of the scale specified in ASTM D291-07. We dropped samples from a hopper placed at a height of 1 m from the bottom of the sample box (Fig. 2) into a sample box with an internal dimension of 150×150×150 mm. The drop test was conducted in the following two ways depending on the purpose.

Fig. 2.

Apparatus for the drop test.

2.1.1. Drop Test of a Briquette Blended into Powder Coal

We first blended only one briquette to investigate in detail the effect of briquette blending on the packing structure of the powder coal portion. The drop test was conducted in three steps. First, a quantity of powder coal filling approximately half the height of the sample box was charged from the hopper (Step 1). Subsequently, a briquette was charged from the hopper, aiming near the top of the pile of the powder coal in the sample box (Step 2). Finally, powder coal was charged from the hopper (Step 3). In this test, the behaviour of the coal during charging at each step was observed with a high-speed camera. Furthermore, after completing each step, we used an X-ray CT system to observe the packing structure of the coal in the sample box.

2.1.2. Drop Test of 20% Briquette Blend

This drop test was conducted by charging the hopper with 4.25 kg (wet basis) of blended coal (8:2 ratio by weight of powder coal to briquettes), simulating the briquette blending ratio in the actual operation. In this test, unlike in Section 2.1.1, the blending of the coal was not divided into steps; instead, the coal was charged into the sample box all at once. Furthermore, the packing structure of the coal in the sample box after the drop test was observed with the X-ray CT system.

2.2. Observation and Analysis by X-ray CT

We used an X-ray CT system (TSX-201 Aquilion LB, Toshiba Medical Systems Corp., Tokyo) to image the coal samples containing briquettes, setting the field of view to 200 to 250 mm, the tube current to 400 mA, the tube voltage to 120 kV, and the imaging pitch to 0.5 to 1.0 mm. By the scanning, CT values are obtained that reflects the density of the scanned objects. When the CT images were converted into 2D images, the CT values were designated to each pixel with brightness according to the CT value. In this study, the brightness range of the obtained X-ray CT images was adjusted to obtain an accuracy of approximately 0.003 t/m³, and then, 8-bit images were extracted and estimated by dividing them into high and low CT values. Under these imaging conditions, the image resolution and bulk density accuracy were 0.391 to 0.488 mm/pixel and 0.003 t/m³, respectively.

Image analysis was performed on the resulting X-ray CT images using the image analysis software ImageJ, WinROOF and Avizo. When applying morphological processing, such as dilation and erosion, in the image analysis, the lattices were analysed as regular hexagonal lattices and the number of connections was set to 4. To obtain the average density in the analysis area, we determined the relationship between the CT values and the density in advance calculated bulk density (BD) (wet, t/m³). Air (0 g/cm³) and water (1 g/cm³) were used for the calibration and their CT values were –1000 and 0, respectively. Based on the information, the relationship between CT value and the density can be expressed as   

$$\rho \left(\text{wet,}\mathrm{\hspace{0.17em} } {\text{t/m}}^{\text{3}}\right)=\frac{{\rho }_{\text{w}}-{\rho }_{\text{a}}}{C{T}_{\text{w}}-C{T}_{\text{a}}}\left(CT-C{T}_w\right)+{\rho }_w=0.001\times CT+1$$(1)

, where ρ and CT represents density and CT value, respectively, and the indices of “w” and “a” means water and air. Based on the equation, the wet basis densities of the objects in each pixel were calculated, and then subtracted the effect of moisture to obtain the dry basis densities.

First, to obtain an overview of the change in the bulk density of the powder coal portion (BD_P) owing to briquette blending, we examined the change in the vertical direction owing to briquette blending. We estimated the average bulk density in a 50 mm × 50 mm horizontal cross-section (xy plane) centred on the briquette (Fig. 3). Here, to eliminate the influence of briquettes on each other, the analysis was focused on the one-briquette drop test.

Fig. 3.

Analysis areas for the evaluation of the change in the bulk density of the powder coal in the coal falling direction. (Online version in color.)

Furthermore, two assumption was made for the packing structure of powder coal portion; i) the packing structure of the powder coal portion would change due to the existence of the briquettes, and ii) the further away from the briquette, the less influence the briquette blend would have on the packing structure of the powder coal. Based on the assumptions, we quantitatively estimated the change in the bulk density of the powder coal portion using the distance w from the briquette as a parameter. Specifically, we estimated the bulk density of the powder coal portion BD_P(w) while excluding from the analysis area regions with a width of w mm traced around each briquette. First, the area without briquette was evaluated (analysis area shown in Fig. 4(b)), and the BD_p(0) was calculated as the mean bulk density of the area. Then, the area w mm away from briquettes was obtained by excluding the regions with a width of w mm traced around each briquette (analysis area shown in Fig. 4(c)), and the mean bulk density of the area was defined as BD_P(w). Finally, the relationship between w and BD_P(w) was examined. For the one-briquette drop test, we analysed (1) vertical sections (xz sections) (Fig. 5(a-1)) and (2) horizontal sections (xy sections) (Fig. 5(a-2)) within an area of 100 mm² centred on the briquette. Whilst, for the 20% blend test, we analysed vertical sections (xz sections) (Fig. 5(b)) within an area of 100 mm² extending up to 25 mm from the interior surface of the sample box.

Fig. 4.

Analysis areas for the evaluation of the bulk density of the powder coal w mm away from briquettes. (a) Original image, (b) analysis area representing where the briquettes are not contained, and (c) analysis area representing w mm away from the briquettes. (Online version in color.)

Fig. 5.

Analysis areas for the evaluation of the bulk density of the powder coal w mm away from briquettes. (a-1), (a-2) Analysis areas for the drop test with powder coal and a briquette, and (b) analysis areas for the drop test with coal charge containing 20 mass% of briquettes. (Online version in color.)

3. Results and Discussion

3.1. Drop Test of a Briquette Blended into Powder Coal

3.1.1. Change in the Bulk Density of the Powder Coal Portion owing to Briquette Blending

The behaviours of the powder coal and a briquette observed using the high-speed camera (a) before the fall of the briquette, (b) immediately after the fall of the briquette (non-steady state), and (c) after the fall of the briquette (steady state) are shown in Fig. 6. We observed that, as soon as the falling briquette landed, the powder coal was scattered by the impact (Fig. 6(b)). Owing to this scattering of powder coal, a crater-like depression centred on the briquette remained after the steady state was reached (Fig. 6(c)).

Fig. 6.

Behaviours of powder coal and a briquette when the briquette fell onto the powder coal.⁷⁾ (a) Before the fall of the briquette (steady state), (b) immediately after the fall of the briquette (non-steady state), and (c) after the fall of the briquette (steady state).

Figure 7 shows the results of using the X-ray CT system to observe the packing structure of the coal (a) after the fall of the briquette; and (b) when the box is filled with the rest of the powder coal after the fall of the briquette. Figure 7(a) shows that the scattering of the powder coal by the impact of the falling briquette produced a gap between the briquette and the powder coal below the briquette. Further, as shown in Fig. 7(b), the gaps formed between the briquette and the powder coal remained and became voids even after the box was filled with the rest of the powder coal. The briquette itself apparently acted as a barrier against the powder coal filling the box from above, thus preventing the gap formed below the briquette from being fully filled.

Fig. 7.

X-ray images of the packing structure of the charged coal after step 2 and step 3. (a) After the fall of the briquette, and (b) after the fall of powder coal onto the charged coal and the briquette.

A closer look at the X-ray CT image showed anisotropy in the local voids around the briquette. The X-ray CT images of the horizontal and vertical sections of the charged coal are shown in Figs. 8 and 9, respectively. Figure 8 shows that a low-density areas are formed below the briquette rather than above it. This is consistent with the aforementioned observations and discussion. Furthermore, Fig. 9 shows that the voids below the briquette are formed at both ends, away from the centre of the briquette (the area circled by the broken line in Fig. 9). As the curvature of the depression formed by the impact of the briquette is smaller than the curvature of the briquette itself, the gap between the powder coal and the briquette produced by the impact of the falling briquette is largest at the ends of the briquette (Fig. 7(a)). Consequently, larger voids are likely to be formed at the ends of the briquette.

Fig. 8.

X-ray images of the charged coal after step 3 (one briquette, horizontal section).

Fig. 9.

X-ray images of the charged coal after step 3 (one briquette, vertical section).

A 3D image of the briquette and the pronounced voids around it is shown in Fig. 10. In the figure, the low-bulk-density areas identified by the binarisation process are averaged with a Gaussian filter to extract the voids, and then, only the areas with the most pronounced voids are shown in blue. It is clear from the 2D cross-sectional image that localised voids are formed at the ends and below the briquette.

Fig. 10.

3D image of the briquette and the voids around the briquette.

Figure 11 shows the results of the quantitative estimation of the change in the bulk density of the powder coal portion BD_P in the vertical direction for the sample shown in Fig. 7(b). First, overall, we observed that the bulk density of the powder coal portion tended to increase as it approached the bottom of the sample box. However, the bulk density of the powder coal portion was lower in the height range containing the briquette than in the regions above and below it (Fig. 11(a)). Furthermore, there was a significant decrease in the bulk density of the powder coal portion in the region around the bottom of the briquette (Fig. 11(b)). The bulk density of the powder coal portion was higher just below the briquette (Fig. 11(c)).

Fig. 11.

Vertical change in the bulk density of the powder coal. (Online version in color.)

Based on the above observations and analysis, it can be concluded that the briquette blending resulted in a decrease in the bulk density of the powder coal portion in the region containing the briquette, and the decrease in the bulk density was particularly significant around the bottom of the briquette. The high-speed camera observations suggest that, after the scattering of the powder coal by the impact of the falling briquette caused the gaps, the briquette itself acted as a barrier against the powder coal filling the box from above, thus preventing the gap formed below the briquette from being fully filled and leaving large voids. Simultaneously, the bulk density of the powder coal portion decreased even above the briquette where no significant voids were observed. This is presumably due to the large difference in size between the briquette and the powder coal, which causes a wall effect between them and hinders the filling of the powder coal around the briquette. It also appears that the impact of the falling briquette compacts the powder coal directly underneath it, causing the voids in the powder coal to be smaller towards the centre of the briquette. Note that the effect of compaction owing to the impact of the falling briquette is small compared with the effect of the low-bulk-density region formed around the briquette.

3.1.2. Quantification of the Width of the Low-bulk-density Region around the Briquette

Subsequently, we quantitatively estimated the extent of the contribution of the reduction in bulk density owing to briquette blending. Figures 12 and 13 show the results of analysing the change in the bulk density of the powder coal portion BD_P(w) in the vertical and horizontal directions, respectively, while excluding from the analysis area the briquette itself and a region with a width of w mm traced around it, based on the methods in Figs. 4, 5(a-1), and 5(a-2). In both analysis areas, the larger the w, the higher is the BD_P, but the change in BD_P decreased once w exceeded a certain value. The absolute values of the bulk density differed between the vertical and horizontal sections. This is attributed to the fact that the analysis range is different between the two, and the fact that the lower part of the sample box, which has a higher bulk density in the vertical section, is included in the vertical section. Furthermore, considering the derivative dBD_P/dw, which expresses the change in the bulk density of the powder coal portion with respect to w, as the width w of the region excluded from the powder coal portion increases, dBD_P/dw decreases and approaches 0. The decrease varies depending on the analysis area; the value of w at which dBD_P/dw = 0 is approximately 10 mm in the vertical section and approximately 6 mm in the horizontal section.

Fig. 12.

BD_P and dBD_P/dw of the bulk density of the powder coal w mm away from the briquette (one briquette, vertical section).

Fig. 13.

BD_P and dBD_P/dw of the bulk density of the powder coal w mm away from the briquette (one briquette, horizontal section).

The larger the w, the higher is the BD_P, which indicates that the bulk density of the powder coal portion is lower near the briquette than away from the briquette. Therefore, it is conceivable that the value of w at which dBD_P/dw = 0 indicates the width of the low-bulk-density region around the briquette. As the value of w for which dBD_P/dw = 0 was larger in the vertical section than in the horizontal section, it can be said that the low-bulk-density region has a wider shape in the vertical direction than in the horizontal direction.

Considering these results together with the observations in Section 3.1.1, the change in the bulk density of the powder coal portion owing to briquette blending can be summarised as follows:

(1) Briquette blending causes a decrease in the bulk density of the powder coal portion around the briquettes.

(2) Powder coal is less likely to fill the area below the briquettes, creating larger voids.

(3) Not only the larger voids, but also low-bulk-density regions due to wall effects seem to exist.

(4) The width of the low-bulk-density region formed around the briquette is wider in the vertical direction than in the horizontal direction.

3.2. Drop Test of 20% Briquette Blend

Figure 14 shows an example of the packing structure after a drop test conducted on a blend of powder coal and briquettes, using a briquette ratio of 20% to simulate the coal charging conditions during actual operation. The figure shows localised low-bulk-density regions formed around the briquettes, such as the regions enclosed by the dashed lines. That is, even when powder coal and briquettes were continuously charged, local voids were formed in the same manner as in Section 3.1 above, which suggests that the same phenomenon occurs in actual furnaces.

Fig. 14.

X-ray image of the charged coal.

Figure 15 shows a 3D image of the pronounced voids around the briquettes, extracted by the same method as that shown in Fig. 10. The image shows localised voids, indicated by blue, formed below and between the briquettes. Similar to Section 3.1, the voids appear to be at locations left unfilled after the powder coal was scattered by the impact of the falling briquette.

Fig. 15.

3D image of the briquettes and the large voids formed around the briquettes.

Figure 16 shows the results of the attempt to quantify the width of the low-bulk-density region around the briquette based on the method shown in Fig. 5(b). Here, only the vertical section was estimated, based on the finding in Section 3.1.2 that the void width is larger in the vertical direction. The results showed that the bulk density BD_P of the powder coal portion tended to increase with the increase in w. On the other hand, unlike the case when only one briquette was added, BD_P did not remain constant. Considering the derivative dBD_P/dw as w increased, dBD_P/dw tended to decrease, but did not reach 0. There were also two inflection points near w = 5 mm and 8 mm.

Fig. 16.

Changes in BD_P and dBD_P/dw of the bulk density of the powder coal w mm away from briquettes (20 mass% briquette, vertical section).⁸⁾

In quantifying the low-bulk-density region around the briquette, the presence of two inflection points for dBD_P/dw suggests that there are two different widths of voids around the briquettes. Considering these analysis results together with the observations in Section 3.1, it appears likely that the low-bulk-density region with a mean width of approximately 8 mm mainly captures the large voids formed underneath the briquettes. On the other hand, the low-bulk-density regions with an average width of approximately 6 mm appear to form around the briquettes regardless of orientation. This is presumably due to the wall effect, in which the different sizes of the powder coal and the briquettes act as a boundary between them, making it difficult for the powder coal to fill the area around the boundary. Venneti et al.⁹⁾ reported that the wall effect was estimated to cause voidage fluctuations up to a distance of approximately 4 to 5 particle diameters away from a wall. The width of the low-bulk-density regions obtained in the analysis, 6 mm, corresponds to 3 to 4 particles from the average size of the powder coal, which is not significantly different from previous findings. The fact that dBD_P/dw did not reach 0 is believed to be due to the complicated packing structure of the blended coal owing to the increase in the briquette ratio.

4. Conclusion

To estimate the change in the bulk density of the powder coal portion owing to briquette blending, we used drop testing followed by observations of the behaviour of the powder coal and briquettes using a high-speed camera; observations of the coal packing structure using an X-ray CT system; and quantitative estimation of the change in the bulk density of the powder coal portion by applying image analysis to the obtained X-ray CT images. The results showed that briquette blending caused localised low-bulk-density regions to form around the briquettes. There were two types of low-bulk-density regions: those formed around the briquettes regardless of orientation (approximately 5 to 6 mm in width) and those formed below the briquettes (approximately 8 to 10 mm in width).

The formation mechanism of each low-bulk-density region is assumed to be as follows.

(1) Low-bulk-density region formed around the briquette

The wall effect was demonstrated for briquettes that were more than 15 times larger than particles composing the powder coal portion.

(2) Low-bulk-density region formed underneath the briquette

The scattering of the powder coal by the impact of the falling briquette causes gaps to form in the powder coal under the briquette. After the powder coal is charged from above, the gaps below the briquette remain as voids because the space below the briquette is shielded and resists being filled in with powder coal.

This study demonstrated that the presence of a low-bulk-density region around the briquette formed by the above two mechanisms causes a reduction in the bulk density of the powder coal when briquette blending is used.

Nomenclature

BD: Bulk density (t/m³)

CT: CT value (-)

CT_a: CT value of air (-)

CT_w: CT value of water (-)

ρ: Denisty (g/cm³)

ρ_w: Denisty of water (g/cm³)

ρ_a: Denisty of air (g/cm³)

BD_P: Bulk density of the powder coal portion (t/m³)

w: Width of the excluded region from the analysis area (mm)

References

• 1)   S.  Nomura,  T.  Arima and  K.  Kato: Fuel, 83 (2004), 1771. https://doi.org/10.1016/j.fuel.2004.03.006
• 2)   S.  Wakuri,  M.  Ohno,  K.  Hosokawa,  K.  Nakagawa,  Y.  Takanohashi,  T.  Ohnishi,  K.  Kushioka and  Y.  Konno: Proc. AIME 45th Ironmaking Conf., Iron & Steel Society, Warrendale, PA, (1986), 303.
• 3)   J.  Echterhoff,  H. J.  Killich and  H. Z.  Kuyumcu: Proc. AIME 50th Ironmaking Conf., Iron & Steel Society, Warrendale, PA, (1991), 185.
• 4)   M.  Yoshinaga,  M.  Sanada and  M.  Utsunomiya: Proc. AIME 35th Ironmaking Conf., Iron & Steel Society, Warrendale, PA, (1976), 256.
• 5)   K.  Okuhara,  T.  Yamaguchi,  Y.  Shimokawa and  M.  Sanada: Coke Circular, 25 (1976), 312 (in Japanese).
• 6)   R. P.  Dias,  J. A.  Teixeira,  M. G.  Mota and  A. I.  Yelshin: Ind. Eng. Chem. Res., 43 (2004), 7912. https://doi.org/10.1021/ie040048b
• 7)   M.  Watanabe,  Y.  Kubota,  K.  Uebo and  S.  Nomura: CAMP-ISIJ, 29 (2016), 140 (in Japanese).
• 8)   M.  Watanabe,  K.  Uebo and  S.  Nomura: CAMP-ISIJ, 28 (2015), 76 (in Japanese).
• 9)   V. M. H.  Govindarao and  G. F.  Froment: Chem. Eng. Sci., 41 (1986), 533. https://doi.org/10.1016/0009-2509(86)87035-X